Association between total cholesterol and depression or depressive symptoms: A systematic review and meta-analysis accounting for comparator comorbidity.

BACKGROUND: The association between total cholesterol (TC) and depression remains uncertain because study populations, depression ascertainment, comparator groups, metabolic comorbidity, and study precision vary substantially. We conducted a systematic review and meta-analysis to quantify group differences in TC and to evaluate the influence of comparator comorbidity, study characteristics, and small-study effects. METHODS: PubMed, Embase, Web of Science, and the Cochrane Library were searched from inception to January 9, 2026. Observational studies reporting TC in participants with depression or depressive symptoms and a comparator group were included. Hedges'g was pooled using random-effects models with Hartung-Knapp-Sidik-Jonkman (HKSJ) inference. We report 95% confidence intervals (CIs), prediction intervals (PIs), and heterogeneity statistics. Prespecified analyses considered reported cardiovascular disease (CVD), diabetes mellitus (DM), depression ascertainment, study design, population characteristics, and potential overlap among database-derived studies. Random-effects meta-regression evaluated associations of effect size with standard error and log-transformed sample size. RESULTS: Forty-four studies involving 491,854 participants were included. Among studies without reported CVD or DM in either group, the pooled SMD was 0.43 (95% CI, 0.05 to 0.81; 95% PI, -1.54 to 2.40; I2 = 99.84%). The association was not statistically significant in studies reporting CVD (SMD = 0.14, 95% CI, -0.12 to 0.41) or DM (SMD = 0.55, 95% CI, -0.17 to 1.27). In studies with CVD-matched controls, the pooled SMD was 0.08 (95% CI, -0.20 to 0.37). Effect size was positively associated with study standard error, whereas its association with log-transformed sample size was not statistically significant. CONCLUSIONS: The pooled average TC difference was positive but should be interpreted cautiously because of extreme heterogeneity, wide prediction intervals, heterogeneous comparator definitions, and possible small-study effects. Current aggregate-data evidence does not establish a consistent or clinically meaningful depression-specific difference in TC or support its use as a depression-specific biomarker. Nonlinear or range-specific associations cannot be excluded.